key: cord-289916-rgvcimk3
authors: Tleyjeh, Imad M.; Kashour, Zakariya; AlDosary, Oweida; Riaz, Muhammad; Tlayjeh, Haytham; Garbati, Musa A.; Tleyjeh, Rana; Al-Mallah, Mouaz H.; Sohail, M. Rizwan; Gerberi, Dana; Bin Abdulhak, Aref A.; Giudicessi, John R.; Ackerman, Michael J.; Kashour, Tarek
title: The Cardiac Toxicity of Chloroquine or Hydroxychloroquine in COVID-19 Patients: A Systematic Review and Meta-regression Analysis
date: 2020-11-02
journal: Mayo Clin Proc Innov Qual Outcomes
DOI: 10.1016/j.mayocpiqo.2020.10.005
sha: 
doc_id: 289916
cord_uid: rgvcimk3

Objective To systematically review the literature and estimate the risk of Chloroquine (CQ) and hydroxychloroquine (HCQ) cardiac toxicity in COVID-19 patients. Methods We searched multiple data sources including PubMed/MEDLINE, Ovid Embase, Ovid EBM Reviews, Scopus, and Web of Science, and medrxiv.org from November 2019 through May 27, 2020. We included studies that enrolled COVID-19 patients treated with CQ or HCQ, with or without azithromycin and reported on cardiac toxicities. We performed a meta-analysis using the arcsine transformation of the different incidences. Results A total of 19 studies with a total of 5652 patients were included. The pooled incidence of TdP arrhythmia or VT or cardiac arrest was 3 per 1000, 95% CI (0-21), I2=96%, 18 studies with 3725 patients. Among 13 studies of 4334 patients, the pooled incidence of discontinuation of CQ or HCQ due to prolonged QTc or arrhythmias was 5%, 95% CI (1-11), I2=98%. The pooled incidence of change in QTc from baseline of ≥ 60 ms or QTc ≥ 500 ms was 9%, 95% CI (3-17), I2=97%. Mean/median age, coronary artery disease, hypertension, diabetes, concomitant QT prolonging medications, ICU care, and severity of illness in the study populations explained between-studies heterogeneity. Conclusions Treatment of COVID-19 patients with CQ or HCQ is associated with a significant risk of drug-induced QT prolongation and relatively higher incidence of TdP/VT/cardiac arrest. Therefore, these agents should not be used routinely in the management of COVID-19 disease. COVID-19 patients who are treated with antimalarials for other indications should be adequately monitored.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that causes the coronavirus disease 2019 (COVID- 19) has spread across the globe claiming hundreds of thousands of lives and causing enormous economic losses. Repurposing of approved drugs for the treatment of COVID-19 was a logical approach before an effective vaccine is available.

Among the drugs that received significant early attention were the antimalarial medications chloroquine (CQ) and hydroxychloroquine (HCQ). CQ and HCQ are weak bases that increase the pH of the intracellular vesicles like endosomes. 1 These changes could affect several stages of viral life cycles from cell entry, viral replication, and viral particle assembly to viral particle release from the host cells. 1 Additionally, these vesicle pH changes interfere with antigen processing and presentation and consequently immune cell activation and production of proinflammatory cytokines. 1 This effect is favorable in the management of autoimmune diseases like systemic lupus erythematosus (SLE) and may have favorable impact in COVID-19 patients with cytokine storm. HCQ was also reported to improve endothelial function and reverse prothrombotic state, 2 which, is relevant to patients with COVID-19 since they manifest significant pulmonary vascular endothelialitis and microvascular thrombosis. 3 The successful demonstration of in vitro antiviral properties of CQ and HCQ against SARS-CoV-2 virus led to initiation of several clinical studies testing the therapeutic potential of CQ and HCQ in COVID-19. 4, 5 The early encouraging experience of CQ therapy in 100 patients from China led to its recommendation by the National Health Commission of China. 6, 7 Another non-randomized study of 20 COVID-19 patients treated with HCQ alone or in combination with J o u r n a l P r e -p r o o f azithromycin in France showed reduced nasopharyngeal viral carrier sate at 6 days after the initiation of treatment. 8 Despite its serious methodological limitations, this report received exceptional attention by media and politicians and triggered a widespread off label use of CQ and HCQ for COVID-19 with subsequent reports of CQ-related deaths. 9

While CQ and HCQ are generally considered safe, QT prolongation and torsade de pointes (TdP) ventricular tachycardia as well as other arrhythmias, myocarditis, and cardiomyopathy have been reported with chronic HCQ use. [10] [11] [12] [13] Recent reports confirmed the increased risk of QT prolongation among COVID-19 patients. 14, 15 The indiscriminate use of the antimalarial medications for the treatment of COVID-19 in the absence of robust clinical evidence for their efficacy, coupled with associated potential harm, calls for rigorously conducted systematic reviews/meta-analyses of the available clinical data to present a clearer picture about their safety and efficacy and provide data-informed view regarding their utility in the treatment of COVID-19. In this study, we set to systematically review the literature regarding the cardiac toxicity of CQ or HCQ in patients with COVID-19.

We followed Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 16 and Meta-Analysis of Observational Studies in Epidemiology (MOOSE) 17 guidelines for reporting systematic review and meta-analysis of observational studies guidelines. Studies that reported electrocardiographic changes and/or cardiac arrhythmias in COVID-19 patients treated with CQ or HCQ with and without azithromycin were included using pre-specified inclusion J o u r n a l P r e -p r o o f criteria as follows: (1) COVID-19 patient population, (2) the study included more than 10 patients receiving either one of the agents, (3) electrocardiographic changes and/or cardiac arrhythmias were reported. In order to avoid introducing non-independence by including patients in the analysis more than once, we included results from the same study with the larger sample size if there were more than one study reporting data on overlapping patient population. 14

The literature was searched by a medical librarian for the concepts of CQ or HCQ combined with COVID-19 on several databases including PubMed/MEDLINE, Ovid Embase, Ovid EBM Reviews, Scopus, and Web of Science. The search strategies were created using a combination of keywords and standardized index terms and were run up to May, 27, 2020 (Supplement 1). We also searched for unpublished manuscripts using the medRxiv, in addition to Google Scholar and the references of eligible studies and review articles. 

Endpoints included the incidence of: (1) change in QTc interval from baseline of ≥ 60 ms, (2) QTc ≥ 500 ms, (3) the composite of endpoint 1 and 2, (4) TdP arrhythmia, or ventricular J o u r n a l P r e -p r o o f tachycardia (VT) or cardiac arrest and (5) discontinuation of treatment due to prolonged QT or arrhythmias.

The number and percentage of subjects experiencing different endpoints were extracted from each study. Due to the very low incidence of TdP and other endpoints (rare events), the arcsine transformation was used to obtain a pooled estimate of the different incidences. For metaanalyses of rare events, this transformation is more appropriate than the commonly used logit transformation, as it can accommodate studies with no observed events, without requiring a continuity correction. [19] [20] [21] Both fixed-effects and random-effects DerSimonian & Laird models with inverse variance method were considered. 22 Statistical tests for heterogeneity were performed using the Cochrane's Q and I-squared statistics, with the assumption of homogeneity being considered invalid for P values < 0.05. 23 When homogeneity could not be assumed, we reported summary estimates and forest plots from the random-effects models.

Meta-regression analyses were used to assess whether the incidence of different endpoints significantly varied by multiple variables specified a priori. These variables were chosen based on risk factors that are known to potentially increase the risk of cardiac arrhythmias (age, sex, CAD, CHF, DM, disease severity, CKD, ICU care, and mortality as another surrogate of disease severity).

We then performed sensitivity analyses by repeating the above analyses after excluding two studies that used high dose CQ or HCQ. Two-tailed p values less than 0.05 were considered to be statistically significant. All statistical analyses were performed using meta and metafor packages in R statistical software, version 3.6.3. 24 J o u r n a l P r e -p r o o f

A total of 19 studies with a total of 5652 patients, including single center and multicenter studies, were included in our systematic review. 14,15,25-41 Figure 1 shows the result of our search strategy (PRISMA flow diagram). Table 1 illustrates the general characteristics of the included studies.

All included studies were of high methodological quality in terms of exposure ascertainment or outcome assessment (Table 2) . Several important cardiovascular adverse events have been observed in our meta-analysis. However, several other events have not been reported consistently by all of the included studies and therefore, we were unable to perform a meta-analysis for all the events. The reported events that we could not meta-analyze were cardiac complications other than TdP, ventricular tachycardia and cardiac arrest. 

TdP tachycardia, ventricular tachycardia and cardiac arrest events were observed in 156 patients in 17 studies with total of 3725 patients. The pooled incidence of these events was 3 per 1000, 95% CI (0-21), I 2 =96% (Figure 2 ). However, only two episodes of TdP tachycardia were reported among 2719 patients. The pooled incidence of discontinuation of CQ or HCQ due to prolonged QTc or arrhythmias was 5%, 95% CI (1-11), I 2 =98% among 4334 patients from 13 studies (Figure 3 ). Among 11 studies with 3127 patients, the pooled incidence of change in QTc from baseline of ≥ 60 ms or QTc ≥ 500 ms was 9%, 95% CI (3-17), I 2 =97% (Figure 4) . In 12 studies of 2008 patients, the pooled incidence of change in QTc from baseline of ≥ 60 ms was 7%, 95% CI (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) , I 2 =94% ( Figure 5 ). Further, the pooled incidence of QTc ≥ 500 ms was 6%, 95% CI (2-12), I 2 =95% from 16 studies with 2317 patients, (Figure 6 ).

For TdP arrhythmia or VT or cardiac arrest, concomitant QT prolonging medications prevalence in included studies explained heterogeneity. The amount of heterogeneity accounted for concomitant QT prolonging medications was 90%. Regarding discontinuation due to prolonged QTc or arrhythmias, Age, CAD, and DM explained heterogeneity. The amount of heterogeneity accounted for by age, CAD and DM was: 40%, 67% and 61%, respectively. For the change in QTc from baseline of ≥ 60 ms, age, CAD, and HTN and ICU admission explained heterogeneity.

The amount of heterogeneity accounted for by age, CAD, HTN and ICU care was: 73%, 82%, 54% and 53%, respectively. On the other hand, heterogeneity for QTc ≥ 500 ms was explained by age, CAD, and HTN and severity of illness. The amount of heterogeneity accounted for by mean/median age, and CAD, and HTN prevalence and severity of illness was: 68%, 65%, 68% and 63%, respectively. Finally, age, CAD, HTN, concomitant QT prolonging medications, ICU care and severity of illness were responsible for observed heterogeneity for the change in QTc from baseline of ≥ 60 ms or QTc ≥ 500 ms. The amount of heterogeneity accounted for mean/median age, and CAD, HTN, concomitant QT prolonging medications and ICU care prevalence and severity of illness was: 80%, 85%, 61%, 58%, 66% and 58%, respectively.

After excluding the 2 studies that used high dose CQ or HCQ, 25,26 we did not observe an important change in pooled estimates or associated heterogeneity.

In this meta-analysis, we systematically examined the risk of QTc prolongation and its associated complications in patients with COVID-19 treated with antimalarial medications CQ and HCQ (CQ/HCQ). The majority of studies included in this analysis used HCQ alone or in combination with azithromycin. Our analysis revealed that treatment with CQ/HCQ was associated with a clinically significant increased risk of QTc prolongation and discontinuation of drug due to this risk. In addition, CQ/HCQ was associated with a clinically significant risk of TdP or VT or cardiac arrest of 3 per 1000 (95% CI 0.0-21).

The incidence of critical QTc prolongation defined as QTc ≥ 500 ms or ∆QTc ≥ 60 ms ranged from 0% to 36%. One of the most remarkable findings is that in the study by Bessière et al, 93% of the studied 40 patients exhibited an increase in QTc prolongation and 36% had critical QTc prolongation. 31 In our pooled analysis, critical QTc prolongation ranged between 6% and 9% with significant heterogeneity among the studies (I 2 of up to 98%). Several factors contributed to the observed heterogeneity were identified by the meta-regression analysis. These include age, hypertension, CAD, ICU admission, DM, use of other QTc prolonging agents, and COVID-19 disease severity. These factors are concordant with the biological explanation for the observed differences, as it is well known that underlying cardiac conditions, comorbidities, and inflammatory states increase the risk of drug induced QTc prolongation. 42, 43 This meta-analysis revealed low but clinically significant risk of combined endpoint of TdP, ventricular tachycardia and cardiac arrest. However, we could not perform a meta-analysis on TdP separately because there were only two reported cases of TdP among 2719 patients from 16 studies (0.073%). The low incidence of TdP is probably an underestimate. A number of factors can explain this low incidence of TdP; most importantly the precautionary discontinuation of the drugs when QTc reaches a certain threshold (QTc > 500 ms or ∆QTc ≥ 60 ms), short duration of therapy, and in certain instances, the therapeutic intervention for long QT using QT shortening agents as reported by Saleh et al for example. 30 Indeed, in our pooled analysis, 5% of patients had their medication discontinued because of QTc prolongation and in one study by Jain et al, 30% of patients had their CQ/HCQ discontinued because of QTc prolongation. 33 Moreover, some TdP cases could have been missed because of underreporting or misclassification. In fact, the two largest studies included in this meta-analysis did not specifically include TdP as a separate endpoint but grouped all arrhythmias under one category.

The study by Rosenberg et al observed arrhythmias in 19.3% and cardiac arrest in 15% of patients and it is quite possible that some of these arrhythmias were TdP or some of the cardiac arrests were preceded by TdP. 36 Nonetheless, the incidence of TdP reported here are consistent with the published data on the drug induced TdP. The risk of developing TdP in association with non-antiarrhythmic drugs is relatively low; for instance, the risk for cisapride, which has been removed from the market, was 0.001%. 44 The risk of TdP with other non-antiarrhythmic drugs is in the range of that was reported with cisapride. The observed incidence of TdP observed in this meta-analysis (0.073%) is 73-fold higher than that of cisapride. It is noteworthy that cisapride and terfenadine (a nonsedating antihistamine) were taken off the market because the risk of TdP even though the risk of TdP associated with their use in absolute terms was very low. This underscores the importance of taking into account the total number of potential lethal events rather than the expressed ratios when assessing the risk drug induced arrhythmias. It is also well known that the highest risk for drug-induced QT prolongation and TdP is associated with class-III antiarrhythmic drugs, which ranges between 1-3% over 1-2 years. 45 The risk of TdP with Sotalol therapy at a low daily dose of 80 mg is only 0.3%. 46 This risk is much higher than the observed risk with CQ and HCQ in this study; however, the estimated risk reported for the antiarrhythmic drugs was over 1 to 2 years of chronic use as opposed to the risk reported here for CQ and HCQ over a very short-term use. In fact, this increases further the concern about the cardiac risk associated with CQ and HCQ treatment in COVID-19 disease.

A number of other cardiac adverse events documented in the included studies were not negligible and include, myocardial injury, acute MI, myocarditis, and others. Notwithstanding, a cause and effect relationship between CQ/HCQ exposure and these complications cannot be J o u r n a l P r e -p r o o f inferred from these studies. However, it is noteworthy to mention that in the study by Borba et al, the incidence of acute cardiac injury was higher in the high dose CQ group in comparison with low dose CQ (50% vs. 31.6%) and the two patients with sustained ventricular tachycardia also occurred in the high dose CQ, which could imply dose-response relationship and probable cause and effect link. 26 . In a recent meta-analysis of HCQ effects on cardiovascular system, ventricular hypertrophy was noted in 22% of patients while heart failure was noted in 26.8% of patients. 48 It is worth noting also that it has been a common practice to use HCQ in combination with azithromycin for COVID-19 during the current pandemic. Azithromycin has been identified as a potential cause of significant serious cardiac arrhythmias through QT prolongation dependent and independent mechanisms and has been linked to increased risk of sudden cardiac death. 49, 50 Hence, the concomitant use of CQ/HCQ and azithromycin or other QT prolonging agents could potentially increase the risk of serious cardiac arrhythmias and death particularly in critically ill patients or those with risk factors for QT prolongation.

Our findings indicate that the cardiac risk imposed by CQ/HCQ use in COVID-19 disease is not trivial and support the need for close monitoring of COVID-19 patients who are treated with CQ/HCQ alone or in combination with azithromycin. Since their efficacy in improving the outcomes of COVID-19 patients is lacking, these agents should be used only in the context of randomized clinical trials given the potential harm that could be associated with their widespread use. This position is supported by the recent FDA statement. 51

Our meta-analysis is the first comprehensive systematic review examining the risk of QT prolongation and its associated adverse events in COVID-19 patients treated with CQ/HCQ. However, like any meta-analysis, it has several limitations. First, the retrospective nature of most of the included studies make them prone to incomplete or missing data. Second, there were variations in the variables collected by individual studies particularly related to reporting QTc parameters and adverse events and significant differences in the patient populations enrolled by these studies. Third, there was marked heterogeneity in our pooled estimates; however, we performed a meta-regression that allowed us to identify contributors to the observed heterogeneity and further determine populations at risk for CQ/HCQ induced QT prolongation, which further strengthens our study and its conclusions.

Our meta-analysis indicates that the treatment of COVID-19 patients with CQ or HCQ alone or in combination with azithromycin is associated with a significant risk of drug-induced QT prolongation. CQ/HCQ use resulted in a relatively higher incidence of TdP as compared to drugs withdrawn from the market for this particular adverse effect. Therefore, these agents should be used only in the context of randomized clinical trials in patients at low risk for drug-induced QT prolongation and with adequate safety monitoring.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. 

A review on possible modes of actions of Chloroquine/ Hydroxychloroquine: Repurposing against SAR-COV-2 (COVID 19) pandemic

Hydroxychloroquine reverses the prothrombotic state in a mouse model of antiphospholipid syndrome: Role of reduced inflammation and endothelial dysfunction

Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in Covid-19

In vitro antiviral activity and projection of optimized dosing design of hydroxychloroquine for the treatment of severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2)

FDA approved drugs with broad anti-coronaviral activity inhibit SARS-CoV-2 in vitro

Breakthrough: Chloroquine phosphate has shown apparent efficacy in treatment of COVID-19 associated pneumonia in clinical studies

Multicenter Collaboration Group of Department of Science and Technology of Guangdong Province and Health Commission of Guangdong Province for chloroquine in the treatment of novel coronavirus pneumonia. Expert consensus on chloroquine the treatment of novel coronavirus pneumonia

Hydroxychloroquine and azithromycin as a treatment of COVID-19: results of an open-label non-randomized clinical trial

Chronic hydroxychloroquine use associated with QT prolongation and refractory ventricular arrhythmia

Hydroxychloroquineinduced cardiotoxicity in a 39-year-old woman with systemic lupus erythematosus and systolic dysfunction

Chloroquine-induced cardiomyopathy: a reversible cause of heart failure

Hydroxychloroquine-induced cardiomyopathy: case report, pathophysiology, diagnosis, and treatment

QT interval prolongation and torsade de pointes in patients with COVID-19 treated with hydroxychloroquine/azithromycin. Heart Rhythm

Risk of QT interval prolongation associated with use of hydroxychloroquine with or without concomitant azithromycin among hospitalized patients testing positive for Coronavirus Disease 2019 (COVID-19)

The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration

Meta-analysis of observational studies in epidemiology: A proposal for reporting

The Newcastle-Ottawa scale (NOS) for assessing the quality of non-randomized studies in meta-analysis

Incidence of atrioventricular block with vasodilator stress SPECT: A meta-analysis

Measures of single arm outcome in meta-analyses of rare events in the presence of competing risks

Why add anything to nothing? The arcsine difference as a measure of treatment effect in meta-analysis with zero cells

Meta-analysis in clinical trials

Measuring inconsistency in metaanalyses

Assessment of QT intervals in a case series of patients with Coronavirus Disease 2019 (COVID-19) infection treated with hydroxychloroquine alone or in combination with azithromycin in an intensive care unit

Hydroxychloroquine and tocilizumab therapy in COVID-19 patients -An observational study. medRxiv

Enhanced ECG monitoring of COVID-19 patients. Heart Rhythm

Early treatment of COVID-19 patients with hydroxychloroquine and azithromycin: A retrospective analysis of 1061 cases in Marseille

No evidence of rapid antiviral clearance or clinical benefit with the combination of hydroxychloroquine and azithromycin in patients with severe COVID-19 infection

Association of treatment With hydroxychloroquine or azithromycin with in-hospital mortality in patients with COVID-19 in New York state

COVID-19 in solid organ transplant recipients: A single-center case series from Spain

COVID-19 in solid organ transplant recipients: Initial report from the US epicenter

Experience with hydroxychloroquine and azithromycin in the COVID-19 pandemic: Implications for QT interval monitoring

Arrhythmic profile and 24-hour QT interval variability in COVID-19 patients treated with hydroxychloroquine and azithromycin

Inpatient use of ambulatory telemetry monitors for COVID-19 patients treated with hydroxychloroquine and/or azithromycin

Institution-wide QT alert system identifies patients with a high risk of mortality

High prevalence of corrected QT interval prolongation in acutely ill patients is associated with mortality: results of the QT in Practice (QTIP) Study

Spontaneous adverse event reports of serious ventricular arrhythmias, QT prolongation, syncope, and sudden death in patients treated with cisapride

Drug-induced long QT syndrome

An update on risk factors for drug-induced arrhythmias

Cardiac complications attributed to chloroquine and hydroxychloroquine: A systematic review of the literature

Azithromycin and the risk of cardiovascular death

Azithromycin causes a novel proarrhythmic syndrome

FDA cautions against use of hydroxychloroquine or chloroquine for COVID-19 outside of the hospital setting or a clinical trial due to risk of heart rhythm problems

The data underlying this article are available in the article and in its online supplementary material.

Dr. Giudicessi reports Equity interest-GlaxoSmithKline, Medtronic, MyoKardia, and Pfizer".However, these relations are not related to this work.Prof. Ackerman reports the following: consultant for Abbott, Audentes Therapeutics, Boston Scientific, Invitae, LQT Therapeutics Medtronic, MyoKardia, and UpToDate. Dr Ackerman and Mayo Clinic are involved in an equity/royalty relationship with AliveCor.However, these relations are not related to this work.Other authors report no conflict of interest.J o u r n a l P r e -p r o o f